Studies in the last two decades have provided a significant new insights into human cancer. Mutations of the retinoblastoma gene (RB) and p53 are involved in about 70 percent of human cancer. These genes, as well as two other tumor suppressor genes, BRCA1 and BRCA2, that are mutated in families predisposed to breast cancer, appear to be also implicated in prostate cancer. We are interested to develop mouse mammary and prostate models of tumorigenesis with an etiology and pathogenesis similar to humans. Due to embryonic lethality and the development of a prevalent type of lethal tumor in young mice, previous models with the above four genes were restricted in their use for cancer research. To circumvent these limitations, we show in this application that the temporal, spatial, and cell-type-specific inactivation of tumor suppressor genes can be achieved in mice using a novel method that places expression of the Cre-loxP recombinase under the regulation of the tetracycline binary system of gene control. This approach has been applied to generate mouse cancer models by conditional inactivation of Rb, p53, Brca1, and Brca2. These mice will be used to: (A) test the specificity of a given tumor suppressor singly or in combination in tumor formation, and to validate the similarity of their tumors to human cancer by pathological and molecular analyses. (B) evaluate the sensitivity of magnetic resonance image analysis for early detection of cancer. (C) investigate environmental factors, such as low dose radiation in tumor formation. (D) test hormone treatments or caloric restriction in preventing tumor formation. (E) evaluate the efficacy of anti-cancer drugs and targeted therapy using tumor suppressor genes in the treatment of breast and prostate cancer in mice. To achieve this goal, four overall specific aims are proposed as follows: (1) To generate mice with tumor suppressor genes, including p53, Rb, Brca1 and Brca2, whose critical exon(s) have integrated flanking loxP (floxed) sites. (2) To establish transgenic mice that demonstrate temporal, spatial and cell-type- specific DNA excision by regulating the expression of the Cre- recombinase with the tetracycline binary system for gene control. (3) To cross mice in Aims 1 and 2 to produce the desired transgenic/floxed mice. (4) To characterize breast and prostate tumor formation and progression in these mice under a variety of settings where tumor suppressor genes, either singly or in combination, are inactivated. In collaboration with other members of the consortium, the involvement of these tumor suppressor genes in other types of cancer could be investigated. In addition, the transgenic mice with regulated Cre-recombinase, either ubiquitously or cell-type- specifically expressed, would be available to other investigators in the consortium to cross with their floxed mice to greatly expand the repertoire of site-specific knock-outs. These valuable mouse models will be extremely useful for the general community of cancer researchers.